Application of ferric anthocyanin chelates as natural blue food colorants in polysaccharide and gelatin based gels

Food Research International 51 (2013) 274–282

Contents lists available at SciVerse ScienceDirect

Food Research International journal homepage: www.elsevier.com/locate/foodres

Application of ferric anthocyanin chelates as natural blue food colorants in polysaccharide and gelatin based gels M. Buchweitz, J. Brauch, R. Carle, D.R. Kammerer ⁎ Hohenheim University, Institute of Food Science and Biotechnology, Chair Plant Foodstuff Technology, Garbenstrasse 25, D-70599 Stuttgart, Germany

a r t i c l e

i n f o

Article history: Received 5 July 2012 Accepted 28 November 2012 Keywords: Blue food colorant Pectin Agar–agar Elderberry Red cabbage Purple carrot Dedicated to Prof. Dr. Dr. h.c. Rainer Wild, Heidelberg, on the occasion of his 70th birthday.

a b s t r a c t Ferric anthocyanin chelates based on elderberry (EB-E) and purple carrot (PC-E) extracts as well as red cabbage juice (RC-J) were applied to different gel matrices to confirm their potential as natural blue food colorants. Blue color evolution and stability during storage at 4 and 20 °C in the dark and at 25 °C under illumination was monitored at two colorant concentrations by determination of lightness (L*), red (a*) and blue (b*) hue values as derived from tristimulus reflectance measurements. Intense blue hues were observed for gels based on gelatin (G) and a blend of agar–agar with amidated pectin (AA/AP). Under each of the storage conditions color stabilities of the gels with PC-E were excellent, being superior to those with added EB-E and RC-J. While room temperature (20 °C) and especially VIS light significantly affected blue color stability in gels dyed with EB-E and RC-J, color of the gels prepared with PC-E was almost fully retained independent of temperature and light exposure. Generally, enhanced stabilities were observed at higher colorant dosages. Gelatin improved color stability significantly compared to AA/AP gels, except for the PC-E colorant, where the difference between both matrices was negligible. Blue color decay in gels, as monitored by increasing b* values, partially deviated from first-order kinetics depending on colorant and storage conditions. Hence, kinetic calculations by exclusive consideration of the b* values were unsuitable for describing and predicting blue color loss. The formation of dairy based gums by substitution of water with yoghurt, buttermilk and milk was only supported by the aid of a gelatin matrix, resulting in pink, violet and slightly blue hues, respectively. In summary, the successful application of ferric anthocyanin chelates to food matrices was demonstrated, confirming their potential as promising natural blue food colorants. © 2012 Elsevier Ltd. All rights reserved.

1. Introduction Due to increasing health awareness, substitution of artificial colorants by their natural counterparts is a major goal of the food, pharmaceutical and cosmetics industry (Stintzing & Carle, 2004). In particular, since the study of McCann et al. (2007), widely known as “Southampton study”, has provided evidence that some azo dyes may be detrimental to children's health, the European Commission imposed labeling of these additives (Directive 1333/2008 (EC)). For replacing red, orange and yellow hues, natural alternatives such as hydrophilic anthocyanins (E 163), betalains (E 162) and lipophilic carotenoids (carotenes, E 160 a–e; xanthophylls, E 161a–j) are available to dye foods (Giusti & Wrolstad, 2003; Henry, 1996; Malien-Aubert, Dangles, & Amiot, 2001; Wissgott & Bortlik, 1996). To avoid E numbers or chemical terms in Abbreviations: ACN, anthocyanin; AA, agar–agar; AP, amidated citrus pectin; DE, degree of esterification (%); DA, degree of amidation (%); GalA, galacturonic acid; ss, soluble solids; aw, water activity; G, gelatin; M, gelatin gel with milk; BM, gelatin gel with buttermilk; Y, gelatin gel with yoghurt; EB, elderberry; RC, red cabbage; PC, purple carrot; J, juice; E, phenolic extract; %, w/w, mass concentration; D, destruction value; L*, lightness; a*, red-green hue; b*, yellow-blue hue; ΔE, color difference. ⁎ Corresponding author. Tel.: +49 711 459 22995; fax: +49 711 459 24110. E-mail address: [email protected] (D.R. Kammerer). 0963-9969/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodres.2012.11.030

the list of ingredients, coloring foodstuffs such as elderberry, red beet or carrot juice concentrates are preferred for dying uncolored products such as candies, confectionary, soft drinks and drugs (Downham & Collins, 2000). In contrast, to achieve blue tones, only a few natural colorants are commercially available. Usually, extracts of Spirulina spp. are applied; however, mainly for coatings, and data regarding color evolution and stability are scarce. Castañeda-Ovando, Pacheco-Hernández, Páez-Hernández, Rodríguez, and Galán-Vidal (2009) suggested intense blue anthocyanin–metal chelates to be an interesting option for natural blue colorants. Surprisingly, no further attempt was undertaken to realize this concept. Blue anthocyanin based colors have been investigated in various flowers. These exceptionally intense and stable hues are due to several stabilizing mechanisms which have been summarized by Yoshida, Mori, and Kondo (2009). In many cases, anthocyanin–metal chelates and additional co-pigmentation with colorless polyphenols in a non-stoichiometric ratio are responsible for the blue hues; however, such fuzzy metal chelates are only stable in their vacuolar matrix, fading away during complex isolation and crystallization. In previous studies we observed a stabilizing effect of pectic structures on the latter preventing complex precipitation in aqueous solutions (Buchweitz, Nagel, Carle, & Kammerer, 2012a). Furthermore, the hydroxylation pattern of the anthocyanin B ring, the presence of uncolored phenolics and

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

further matrix compounds as well as structural prerequisites of pectic compounds to achieve blue color evolution and enhanced stability were thoroughly investigated in model solutions (Buchweitz, Carle, & Kammerer, 2012b; Buchweitz et al., 2012a). In small scale model experiments promising blue hues, storage and heat stabilities were demonstrated for ferric anthocyanin chelates (Buchweitz, Brauch, Carle & Kammerer, in press). Therefore, the objective of the present study was to transfer these results to real foods such as edible gels, demonstrating the applicability of natural anthocyanin-based blue pigments in food. Colorants based on extracts of elderberry (Sambucus nigra L., EB-E) and purple carrot (Daucus carota L. ssp. sativus var. atrorubens Alef., PC-E) as well as red cabbage (Brassica oleracea L. ssp. capitata f. rubra) juice (RC-J), the latter even enabling its labeling as coloring foodstuff, should be developed and added to protein and polysaccharide based gels under conditions commonly applied in industrial practice. Color evolution and impact of storage conditions (temperature and light) on color stability should be thoroughly evaluated. 2. Materials and methods 2.1. Anthocyanin solutions For preparing the phenolic extracts (E), concentrates of elderberry (EB; Exberry GNT Europe, Aachen, Germany) and purple carrot (PC; Wild, Heidelberg, Germany) were diluted with water and applied to solid phase extraction on a polystyrene adsorber resin (Amberlite FPX66, Rohm & Haas, Philadelphia, PA, USA) following the procedure described previously (Buchweitz et al., in press). Red cabbage juice (RC-J) was prepared by dilution of a concentrate (Exberry GNT Europe, Aachen, Germany) with water. All stock solutions were adjusted to total anthocyanin concentrations of 2.01 ± 0.02· 10 −3 mol/L and stored at −18 °C until further use. Demineralized water was used throughout. A comprehensive chemical characterization of these anthocyanin containing stock solutions, including sugars and acids as well as HPLC chromatograms and polyphenol assignment is provided as supplementary material (Data S1).

275

concentrations of AP, anthocyanins and ferric ions amounted to 0.74% (w/w), 2.11 and 6.32 · 10−4 mol/L, respectively, and a molar anthocyanin to ferric ion ratio (ACN:Fe3+) of 1:3 was obtained. After adjusting the pH value to 5.0 and continuous stirring over night, aliquots of the colorant were filled in polyethylene bottles and stored at −18 °C until further use. Each colorant was prepared in duplicate. 2.5. Model gels 2.5.1. Gels prepared from a commercial pectin formulation (matrix CS 025-B) 25 g of pectin formulation (Pektin Amid CS 025-B) was blended with 100 g sucrose and 1.8 g potassium sorbate. This mixture was stirred into 220 g water and boiled while stirring until the pectin was completely dissolved. Glucose syrup (475 g) and sucrose (260 g) were added, and the solution was cooked to a final soluble solid content (ss) of 78.0 ± 0.1% (measured by refractometer RX-5000, Atago, Japan) as recommended for jelly production by the supplier of the pectin formulation. The final concentration of pectin formulation amounted to 2.5%, and the pH value was 4.5. Subsequently, to each portion of 200 g, 60 mL of colorant together with 20 mL water and 80 mL of colorant for colorant dosages 1 and 2, respectively, were added at 80 °C. 2.5.2. Gels prepared from amidated pectin (matrix AP) The procedure was analogous to Section 2.5.1 using amidated pectin AP instead of the pectin formulation (Pektin Amid CS 025-B). 2.5.3. Gels prepared from agar–agar and amidated pectin (matrix AA/AP) Potassium sorbate (1.2 g) was dissolved in 400 mL water, and 20 g agar–agar (AA, 20%, w/w) was added under stirring. The solution was boiled for 4 min, and 300 g sucrose blended with 2 g amidated pectin (AP) was added. The mixture was cooked to a final ss content of 49%. The concentrations of AA and AP were 0.6 and 0.3%, respectively, and a pH value of 5.5 was determined. Subsequently, to each portion of 200 g, 60 mL colorant and 20 mL water (colorant dosage 1) and 80 mL (colorant dosage 2) were added at 80 °C. The ss content and pH value of the blue gels were 35.7 ± 0.1% and 5.3 ±0.0, respectively.

2.2. Hydrocolloids Low methoxylated amidated citrus pectin (AP; degree of esterification (DE) 27.4%, degree of amidation (DA) 22.8%, content of galacturonic acid (GalA) 93.5%) and a commercial pectin formulation (Pektin Amid CS 025-B) based on amidated pectin (DE 25.8%, DA 20.1%), recommended for gum and jelly products, were donated by Herbstreith & Fox (Neuenbürg, Germany). Commercially available agar–agar (AA; 20% in maltodextrin) and gelatin (G) were from Ruf Lebensmittelwerk (Quakenbrück, Germany).

2.5.4. Gels prepared from gelatin (matrix G) Potassium sorbate (1.25 g) was dissolved in 138 mL water, and 575 g sucrose was added with subsequent heating of the mixture to 100 °C. Gelatin (25 g) was soaked for 5 min in cold water, drained and dissolved in the hot sucrose solution without boiling. The mixture was cooked to a final ss content of 70% (3.4% gelatin, w/w) at a pH value of 5.7. Subsequently, to each portion of 200 g, 40 mL colorant with 20 mL water and 60 mL colorant for colorant dosages 1 and 2, respectively, were added at 80 °C. The ss content and pH value of the blue gels were 58.5 ± 0.1% and 5.7 ± 0.1, respectively.

2.3. Miscellaneous Potassium sorbate p.a. was purchased from Merck (Darmstadt, Germany) and glucose syrup from Schließmann (Schwäbisch Hall, Germany). Crystalline sucrose was from Südzucker (Mannheim, Germany). Cow's milk (M; 1.5% fat), buttermilk (BM, b 1% fat) and yoghurt (Y; 0.1% fat) were from Milbona (Ravensburg, Germany) and purchased on local markets. 2.4. Preparation of the colorant Amidated pectin (AP) was suspended in water (1%, w/w) and shaken over night. 224 mL of this solution was mixed with 32 mL of the respective anthocyanin stock solution (2· 10 −3 mol/L). 48 mL of aqueous ferric stock solution (4· 10 −3 mol/L FeCl3 · 6H2O, Fluka, Buchs, Switzerland) was added slowly. The pH value of both stock solutions was adjusted to pH 3.3 immediately before mixing. The final

2.5.5. Gelatin based gels prepared with dairy products (matrices M, BM, Y) The basic gelatin gel (2.5.4.) was modified by replacing water with the respective amount of milk (M, pH 6.6), buttermilk (BM, pH 4.6) and yoghurt (Y, pH 4.4). To each portion of 200 g, 60 mL of the colorant based on red cabbage juice (RC-J) and 20 mL water were added (colorant dosage 1) and pH values of 6.1, 5.1 and 4.9 were determined, respectively. Additionally, milk-containing gels (M) with the addition of 80 mL RC-J colorant per 200 g (colorant dosage 2) were prepared. All gels were prepared in duplicate (A and B samples). Aliquots of ~ 12 g were filled in Petri dishes (35/10 mm, Greiner bio-one, Frickenhausen, Germany), cooled to room temperature, closed and sealed with Parafilm®. Uncolored gels (blanks) were prepared similarly by adding water instead of colorant. All samples were stored for 48 h at 20 ± 0.2 °C in the dark to establish equilibrium (t = 0).

276

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

Subsequently, samples were stored under the conditions specified in Section 2.6.

3. Results and discussion 3.1. General information

2.6. Storage conditions For shelf life experiments, 10 Petri dishes each for A and B samples of each model gel were stored at 20 ± 2 °C in the dark over two months. Further model gels containing PC-E and RC-J were prepared, and 6 Petri dishes each for A and B samples were stored at 4 ± 2 °C in the dark and at 25 ± 0.5 °C under illumination with visible light (VIS) in a climatic test cabinet (fluorescent light, Osram 35 W/640; Binder, Tuttlingen, Germany). To guarantee uniform light exposure to each sample, Petri dishes were placed exactly in front of a lamp in a distance of 20 cm. Color measurements were performed once and twice a week for storage at 20 °C and 25 °C + VIS, respectively, and every 3 weeks for storage at 4 °C. 2.7. Color analysis and evaluation of pigment degradation For color analyses of the model gels a Minolta Chromameter CR-300 (Konica Minolta, Osaka, Japan), calibrated with a standard white tile (L* = 97.43, a* = 0.01, b* = 1.64), was used. The Petri dishes were placed on a white paper and measured in a dark box. This setup was needed to prevent environmental effects on L*a*b* determination, especially of the slightly transparent AA/AP gels. L*, a* and b* values were determined for A and B samples independently by threefold measurement of 10 or 6 Petri dishes depending on storage condition (Section 2.6). Mean values and standard deviations were calculated from the mean values obtained for A and B. Total color differences (ΔE*) were calculated according to Eq. (1) between different model gels at t = 0 and after storage for 12, 36 and 54 days enabling comparison between the different storage conditions. h
2
2
2 i1=2 ΔE
¼ ΔL
þ Δa
þ Δb


ð1Þ

In model solutions color loss followed pseudo-first-order reaction kinetics (Buchweitz et al., in press). The decay of negative b* values was assumed to indicate blue color loss, and therefore, the natural logarithm of the ratio (b*)t/(b*)0 was plotted against storage time t (Eq. (2)), and the slope of the graph was equated with − k, from which half-life values were calculated (Eq. (3)). 

  ln b
t = b
0 ¼ −k  t

ð2Þ

t1 = ¼ ln2=k

ð3Þ

2

(b*)t, b* value at time t; (b*)0, initial b* value; k, slope and degradation rate constant; t, time [days]. Additionally, destruction values (D values) indicating the time required for 90% color loss were calculated as a further indicator of color stability (Eq. (4); Reyes & Cisneros-Zevallos, 2007). D ¼ ln10=k

ð4Þ

2.8. Statistical analysis Statistical analyses were conducted using the Tukey test (JMP 6 software, SAS, Cary, NC, USA). Data were compared using the least significant difference (LSD) test (α = 0.05).

3.1.1. Procedure for colorant manufacturing Based on the promising results regarding color evolution and stability gained in model solutions at a micro scale (Buchweitz et al., in press), extracts of elderberry (EB-E) and purple carrot (PC-E) were chosen for colorant preparation. Furthermore, despite their poor stability, ferric anthocyanin chelates derived from red cabbage juice (RC-J) were also included due to their outstanding blue hue, broadest pH range stability, and low amounts of ferric ions needed for producing blue hues. Preliminary tests pointed out, that ferric ion concentrations exceeding a molar ACN:Fe3+ ratio of 1:1 and 1:2 as evaluated in model solutions (Buchweitz et al., in press) were necessary to compensate interfering interactions with the gel matrix (data not shown). Therefore, the colorants were prepared with a molar ratio of anthocyanins to ferric ions of 1:3. Similar results regarding chelate formation and stabilization were previously reported for amidated and high methoxylated pectin (Buchweitz et al., in press). Nevertheless, amidated pectin was selected for chelate stabilization in gels due to its good gelling properties already at slightly acidic pH (Endress, Mattes, & Norz, 2009). The scaling up of colorant preparation developed previously in micro scale experiments and replacement of acetate buffer by water were challenging. The pH during preparation turned out to be a key factor to prevent pectin precipitation (as a consequence of enhanced ferric ion–pectin interactions), formation of non-reactive ferric hydroxo complexes and colorless hemiacetal and chalcone forms of the anthocyanins, respectively (Dangles, Elhabiri, & Brouillard, 1994; Elhabiri, Figueiredo, Toki, Saito, & Brouillard, 1997; Endress et al., 2009). 3.1.2. Gel matrices appropriate for coloration with ferric anthocyanin chelates With respect to their matrix, anthocyanin ferric chelates are known to require some basic conditions. Apart from a pH value of 4.5 or even higher being mandatory (Bayer, Egeter, Fink, Nether, & Wegmann, 1966; Dangles et al., 1994; Elhabiri et al., 1997), unesterified carboxylic groups were found to be detrimental to color evolution. The formation of the blue ferric anthocyanin chelates was inhibited due to ionic interactions of ferric ions causing chain aggregation, and hence polysaccharide precipitation (Endress et al., 2009; Jorge & Chagas, 1988). Edible acids such as citric or lactic acids annihilated blue color evolution due to their acidic and iron chelating properties (Buchweitz et al., in press). Consequently, chelating acids had to be avoided and basic formulations based on high methoxylated pectin were inappropriate due to their sugar and acid based gelling mechanism, requiring pH values ≤3.5 (Thibault & Ralet, 2003). As a result, and in consideration of the state-of-the-art of industrial jelly production, amidated pectin (AP), agar–agar (AA) and gelatin (G) were selected for preparing gel matrices. The presence of amide functions in low methoxylated pectin molecules is known to reduce calcium sensitivity. Therefore, amidated pectins can be applied to formulations where non-amidated low methoxylated (LM) pectins will not form gels or even precipitate due to insufficient or excessive calcium concentrations or neutral pH values (Endress et al., 2009). In agar–agar the amount of ionic agaropectin structures, and hence the amount of sulfate and free carboxylic groups, strongly depends on the extraction procedure (Nussinovitch, 1997). Therefore, commercial agar–agar preparations have to be tested prior to their use in gels dyed with ferric anthocyanin chelates. The application of ferric anthocyanin chelates to a commercial pectin preparation recommended for jelly production by the supplier (matrix CS 025-B) failed, although the formulation was based on amidated pectin. According to the supplier's information, added citrate, tartrate and phosphate salts facilitate handling to achieve optimal texture independent of pH value and mineral content of the ingredients (e.g. fruits). However, these retarding agents annihilated chelate formation, and

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

only slightly pink colored gels were obtained. Substitution of the CS 025-B formulation by the amidated pectin AP (also used for preparing the colorant) was not successful. Besides difficult manufacturing, textural properties were unsatisfactory and could not be improved by higher pectin dosages (2.5–4.5%, w/w). However, combining agar– agar and amidated pectin (matrix AA/AP) in a blend enabled convenient manufacture, good gelling properties and optimal color evolution. The addition of pectin (0.3%, w/w) was essential to improve gelation, since gels only made of agar–agar showed brittle fracture and tended towards syneresis. In contrast, the preparation of gelatin based gels (matrix G) was without any difficulty. Therefore, complete substitution of water by dairy products such as milk, buttermilk and yoghurt was only feasible for this matrix without affecting the manufacturing process and texture. Contrary, gels prepared with dairy products based on AA/AP matrix rapidly disintegrated after colorant addition, thus hindering color measurements. Presumably, due to additional calcium ions supplied by the dairy products the enhanced concentration of cations may have caused chain aggregation of the pectin and agar–agar molecules. However, to achieve satisfactory color intensities, increased colorant dosages of RC-J were required for gelatin gels prepared with dairy products compared to gelatin gels solely containing water. 3.2. Impact of gel matrix, colorant dosage and pigment source on color evolution The gels markedly differed in their color depending on the gel matrix, the pigment source of the colorant and colorant dosage, respectively. Gels dyed with PC-E and EB-E at the lower dosages (40 and 60 mL per 200 g G and AA/AP, respectively) displayed blue hues with a violet tint. This was indicated by positive a* values (red hue) in the range of 0.07–0.68 (Fig. 1). The gels containing the higher colorant dosage (60 and 80 mL per 200 g) displayed a pure cobalt blue, characterized by negative a* values (− 0.10 b a* b − 0.82). Due to additional interaction of ferric ions with the gel matrix, enhanced colorant

10

277

dosages (containing more ferric ions) were needed to improve the blue color hue. In contrast, for RC-J a* values were less influenced by the colorant dosages which may indicate a stronger anthocyanin– ferric ion interaction. The polysaccharide matrix (AA/AP) showed a stronger affinity to ferric ions than the protein (G) matrix. For this reason, generally higher colorant dosages for AA/AP gels were necessary to obtain acceptable blue colors. The b* values below zero, indicating blue hues, were more negative for EB-E (−3.59 b b* b −5.58) than for PC-E (−2.40 b b* b −4.12) at comparable dosages and matrices. Appealing bright blue colors (gentian-blue) with strongly negative b* values (− 5.85 b b* b − 9.35) were only observed for gels containing RC-J. These findings about color evolution in gels confirmed previous observations in simple micro scale model solutions (Buchweitz et al., in press). In particular, the evolution of brilliant blue hues of ferric anthocyanin chelates as observed in our previous study was verified when applying red cabbage juice to various matrices at varying dosages. In general, compared to gels based on gelatin, AA/AP gels displayed stronger positive a* values. This effect was already observed in the uncolored matrices (blanks); however, stronger interactions between the ferric ions and the sulfates and/or free carboxylic groups of the polysaccharide molecules hindered unflawed formation of ferric anthocyanin chelates. With the exception of gels dyed with RC-J(2), differences in lightness (L*) between the two gel matrices were insignificant (Table 1A). However, due to the turbidity of gelatin based gels the blue colors appeared to be deeper and more intense under daylight than for the transparent AA/AP gels. Higher colorant dosages resulted in more intensely colored gels exhibiting lower L* values. Whereas L* values between samples prepared with the two dosages differed marginally for the PC-E colorant (ΔL* = 0.6), significant differences of 1.6 (G), 1.0 (AA/AP), 2.7 (G), and 1.8 (AA/AP) were observed for EB-E and RC-J, respectively. Similarly, total color differences (ΔE*(1) − (2)) between the colorant dosages were highest for RC-J, followed by EB-E and PC-E (Table 1B).

+ b*

blank

blank

+ a*

- a*

-3

3 PC-E(2)

PC-E(2) EB-E(2) EB-E(2)

PC-E(1) PC-E(1)

EB-E(1) EB-E(1)

RC-J(2) RC-J(1)

RC-J-BM(1) RC-J-M(1)

RC-J(2)

RC-J-M(2)

RC-J(1)

-10 G

AA/AP

- b*

G with dairy products;

G, gelatin based gel; AA/AP, agar-agar / amidated pectin based gel; blank, model gel without colorant addition; (1), lower amount of colorant; (2), higher amount (+20 mL) of colorant; M, milk; BM, buttermilk; Y, yoghurt; EB-E, elderberry extract; RC-J, red cabbage juice; PC-E, purple carrot extract Fig. 1. Initial red (a*) and blue (b*) hues of gelatin (G) and agar–agar/amidated pectin (AA/AP) based gels.

RC-J-Y(1)

278

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

However, since ΔE* is a sum parameter, differences in lightness are equal to differences in color hue (a* and b*). For the blue gels, the results of color measurements deviated from visual perception. Differences in color hue, in particular in a* values, became more apparent than comparable changes in lightness. Total color differences (ΔE*(G) − (AA/AP)) between the gel matrices were independent of the colorant dosage and most pronounced for RC-J with ΔE*(G) − (AA/AP) values of 2.4. The threshold value for color differences noticeable by the human eye is ΔE* = 1 (Gonnet, 1998). Therefore, differences among the matrices should be negligible for EB-E (ΔE(G)−(Ag/A) ~ 0.8) and PC-E (ΔE*(G)−(AA/AP) = 0.5) (Table 1C), however, probably due to the aforementioned diverse transparency of the matrices, the colored gels appeared significantly different under daylight. 3.3. Impact of dairy products on color evolution Significant differences in color evolution were observed for gelatin based gels prepared with dairy products instead of water which were dyed with RC-J colorant. Gels containing customary milk displayed light blue colors with minor a* values (0.20b a*b 0.29) and pronounced negative b* values in the range of −7.75 to −8.73 (Fig. 1). With Table 1 Initial values of lightness (L*, A), and total color differences (ΔE*) between gels produced with two colorant dosages (B) and comprising different matrices (C) dyed with ferric anthocyanin chelates. A L* Additive

– – – – – – – M M BM Y

Colorant

Gel matrix

Blank EB-E(1) EB-E(2) RC-J(1) RC-J(2) PC-E(1) PC-E(2) RC-J(1) RC-J(2) RC-J(1) RC-J(1)

G

AA/AP

45.8 ± 6.21 27.6 ± 0.5f 26.0 ± 0.3ij 29.5 ± 0.le 26.8 ± 0.1gh 26.1 ± 0.3hij 25.6 ± 0.1j 39.7 ± 0.0b 38.2 ± 0.3c 35.6 ± 0.0d 41.1 ± 0.2a

56.3 ± 371 27.3 ± 0.1fg 26.3 ± 0.0hi 29.6 ± 0.1e 27.8 ± 0.1f 26.4 ± 0.1hi 25.8 ± 0.1ij – – – –

B Gel matrix

Colorant

ΔE*(2)−(1)

G2

EB-E RC-J CA-E EB-E RC-J PC-E

2.2 ± 0.1b 3.0 ± 0.2a 1.8 ± 0.2b 2.0 ± 0.2b 2.3 ± 0.2ab 1.7 ± 0.1b

AA/AP3

C Colorant

ΔE*(AA/AP−G)

EB-E(1) EB-E(2) RC-J(1) RC-J(2) PC-E(1) PC-E(2)

0.9 ± 0.1b 0.7 ± 0.0cb 2.4 ± 0.2a 2.4 ± 0.0a 0.5 ± 0.1c 0.5 ± 0.1c

Mean values ± standard deviation (n = 10); significant differences between values in the same table are indicated by different letters (P b 0.05); L*, lightness; G, gelatin based gel; AA/AP, agar–agar/amidated pectin based gel; ΔE*(2)−(1), total color differences between the two colorant dosages; ΔE*(AA/AP−G), total color differences between the matrices; blank, model gel without addition of colorant; 240 and 60 mL colorant per 200 g; 360 and 80 mL colorant per 200 g; (1), lower colorant dosage; (2), higher colorant dosage; M, milk; BM, buttermilk; Y, yoghurt; EB-E, elderberry extract; RC-J, red cabbage juice; PC-E, purple carrot extract. 1 Variations due to higher impact of unequally distributed turbidity and air bubbles in uncolored gels; therefore, blank gels were not included in statistical analyses.

buttermilk and yoghurt the gels exhibited violet (a* = 1.28) and pink (a* = 2.78) hues, respectively. This negative effect on blue color evolution was attributed to the presence of lactic acid. Buttermilk and yoghurt contain ~600 and ~1460 mg/100 g of lactic acid, respectively (Souci, Fachmann, Kraut), which annihilated the formation of blue ferric anthocyanin chelates due to the complexation of ferric ions (Buchweitz et al., in press). Generally, the white background color of the dairy products produced higher L* values (Table 1). However, the difference between the two colorant dosages of RC-J in these gels was notably smaller compared to the gelatin based gels produced with water. 3.4. Impact of the gel matrix and colorant on color stability The blue gels based on gelatin and agar–agar/amidated pectin were stored at 20 °C in the dark. During storage for 54 days, the colors changed from blue to gray and yellowish green for gels containing EB-E and RC-J, respectively, while the gels dyed with PC-E colorant still retained their blue hues; however, displaying a gray shade. For all gels, lightness increased slightly (G; 0.1 b ΔL* b 1.2; AA/AP, − 0.2 b ΔL* b 3.2) (Figs. 2 and 3). As expected, a* values were constant or decreased marginally (G, − 0.82 b Δa* b − 0.44; AA/AP, − 1.14 b Δa* b − 0.70). The negative b* values, indicative of blue hues, increased steadily during storage reflecting blue hue loss. After 54 days, differences (Δb*) of 1.02–4.31 and 1.34–7.67 were observed for G and AA/AP gels, respectively. Total color differences (ΔE*) were mainly due to the decrease of b* values. Therefore, the color losses considering different pigment sources (EB-E, RC-J, and PC-E), colorant dosages ((1) and (2)) and gel matrices (G and AA/AP) as expressed by ΔE* showed the same trends as those based on Δb* values (data not shown). Regarding the pigment sources, blue color stability decreased in the following order: PC-E> EB-E > RC-E (Table 2). Thus, the results found for the gels confirmed previous findings for blue color stability in liquid formulations (Buchweitz et al., in press). During storage over 54 days, color stability of the RC-J based colorant was significantly better for gels containing G than for those made from AA/AP with ΔE* values of 2.7–4.5 and 6.6–8.4, respectively. For gels containing EB-E, differences between the matrices were less obvious with ΔE* values amounting to 2.2–3.0 and 2.5–3.7 for G and AA/AP, respectively. The outstanding stability of the PC-E colorant (ΔE = 1–2) was almost independent of the gel matrix. Therefore, the stabilizing effect of gelatin was more pronounced for the less stable blue chelates of anthocyanins from RC-J and EB-E than for the very stable chelates of PC-E pigments. If this effect is actually due to the proteinaceous nature or more likely to the slightly higher pH value or the markedly enhanced sugar content (G: pH 5.7, ss content 58.5%; AA/AP: pH 5.3, ss content 35.7%), reducing water activity (aw), remains unknown. Generally, higher colorant dosages went along with enhanced color stabilities for each of the gel matrices and pigment sources. In previous investigations of anthocyanin interactions with hydrocolloids, enhanced stability of the red hue (a* value) in gels (pH 3.3, 70% ss content) containing gelatin and black current concentrate compared to gels prepared with agar–agar was observed (Hubbermann, Heins, Stöckmann, & Schwarz, 2006). However, best stabilities were observed in pectin, unfortunately of undefined DE. Additionally, Maier, Fromm, Schieber, Kammerer, and Carle (2009) observed a stabilizing effect of pectin (DE 56–60%) on anthocyanins in gels enriched with grape pomace extract when stored under illumination. In this case anthocyanin retention was significantly improved compared to gelatin gels, nevertheless, differences between the two matrices during storage without exposition to light at 6 and 20 °C were marginal. 3.5. Impact of storage conditions Due to their highly desired blue color hues and most excellent stabilities at room temperature, gels dyed with PC-E at the elevated dosage

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

279 4

60

2

50

0

40

-2

30

-4

20

-6

10

-8

L*

a*,

b*

70

0 t [d] T [°C]

-10

0 12 36 54 12 36 54 12 36 54 0 12 36 54 12 36 54 12 36 54 0 12 36 54 12 36 54 12 36 54

4

20

25 + VIS

4

20

RC-J(1) ΔL* (0-54) Δa* (0-54) Δ b* (0-54)

c

-3.53 (±0.10) a-e -0.74 (±0.07) defg 1.65 (±0.07)

a

1.19 (±0.35) abc -0.52 (±0.05) c 4.31 (±0.19)

25 + VIS

4

20

RC-J(2) 1

3.44 (±0.32) 1 -0.05 (±0.05) 1 7.86 (±0.56)

c

-3.60 (±0.52) a-f -0.82 (±0.06) efg 1.45 (±0.04)

ab

0.43 (±0.01) a-f -0.82 (±0.05) de 2.51 (±0.12)

25 + VIS

PC-E(2) 1

2.04 (±0.21) 1 -0.57 (±0.02) 1 4.27 (±0.37)

c

-3.24 (±0.28) a-d -0.71 (±0.07) gh 0.99 (±0.04)

b

0.09 (±0.24) a -0.44 (±0.23) gh 1.02 (±0.09)

b

-0.08 (±0.03) a-d -0.59 (±0.05) fg 1.30 (±0.01)

mean values ± standard deviation; significant differences between values in the same line are indicated by different letters (P < 0.05); 1 differences after 36 days of storage and not included in statistical analyses; L*, lightness; a*, red-green hue; b*, yellow-blue hue; (1), lower colorant dosage; (2), higher colorant dosage; RC-J, red cabbage juice; PC-E, purple carrot extract

Fig. 2. Changes of lightness (ΔL*) and color hues (Δa*, Δb*) in blue gelatin based gels (G) after storage for 12, 36 and 54 days at 4 °C, 20 °C and under illumination at 25 °C.

mean values ± standard deviation; significant differences between values in the same line are indicated by different letters 1 (P < 0.05); differences after 36 days of storage and not included in statistical analyses; L*, lightness; a*, red-green hue; b*, yellow-blue hue; (1), lower colorant dosage; (2), higher colorant dosage; RC-J, red cabbage juice; PC-E, purple carrot extract Fig. 3. Changes of lightness (ΔL*) and color hues (Δa*, Δb*) in blue agar–agar/amidated pectin (AA/AP) based gels after storage for 12, 36 and 54 days at 4 °C, 20 °C and under illumination at 25 °C.

280

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

Table 2 Total color differences (ΔE*) of blue gels after storage for 12, 36 and 54 days at 4 °C, 20 °C and under illumination at 25 °C. Storage conditions

Colorant

ΔE* G

4 °C

20 °C

25 °C + VIS

Blank EB-E(1)1 EB-E(2)1 RC-J(1) RC-J(2) PC-E(1)1 PC-E(2) Blank EB-E(1) EB-E(2) RC-J(1) RC-J(2) PC-E(1) PC-E(2) Blank EB-E(1)1 EB-E(2)1 RC-J(1) RC-J(2) PC-E(1)1 PC-E(2)

AA/AP

12 days

36 days

54 days

12 days

36 days

54 days

0.4 ± 0.12 – – 0.5 ± 0.0c 0.5 ± 0.lc – 0.2 ± 0.0c 1.2 ± 0.82 0.5 ± 0.1c 0.3 ± 0.1c 1.1 ± 0.3bc 0.6 ± 0.0c 0.2 ± 0.1c 0.2 ± 0.0c – – – 4.0 ± 0.7a 1.9 ± 0.2b – 0.4 ± 0.0c

1.5 ± 0.12 – – 1.4 ± 0.0def 0.8 ± 0.1g – 0.7 ± 0.0g 1.9 ± 0.92 1.7 ± 0.0de 0.8 ± 0.2g 2.8 ± 0.1c 1.8 ± 0.0d 0.6 ± 0.1g 0.8 ± 0.1fg – – – 8.6 ± 0.4a 4.8 ± 0.2b

1.7 ± 0.22 – – 4.1 ± 0.0ab3 4.0 ± 0.4ab3 – 3.5 ± 0.3bc3 2.4 ± 1.82 3.0 ± 0.lcd 2.2 ± 0.le 4.5 ± 0.3a 2.7 ± 0.1de 2.3 ± 0.2de 1.1 ± 0.0f – – – nd nd

1.1 ± 0.0efg

1.4 ± 0.0f

0.2 ± 0.22 – – 0.4 ± 0.1de 0.2 ± 0.0e – 0.4 ± 0.lde 0.3 ± 0.02 0.9 ± 0.0d 0.6 ± 0.3de 1.8 ± 0.1c 1.5 ± 0.3c 0.6 ± 0.0de 0.6 ± 0.0de – – – 6.8 ± 0.la 4.3 ± 0.2b – 0.4 ± 0.0de

3.8 ± 0.52 – – 1.6 ± 0.1ef 1.3 ± 0.1f – 0.8 ± 0.3f 1.4 ± 1.02 2.5 ± 0.le 1.5 ± 0.3ef 6.0 ± 0.2c 4.8 ± 0.7d 1.4 ± 0.1f 1.0 ± 0.3f – – – 11.9 ± 0.3a 8.3 ± 0.2b – 1.0 ± 0.0f

2.7 ± 1.12 – – 4.3 ± 0.3c3 3.9 ± 0.1cd3 – 3.3 ± 0.3cde3 1.0 ± 0.12 3.7 ± 0.1cd 2.5 ± 0.2def 8.4 ± 0.3a 6.6 ± 1.0b 2.2 ± 0.lef 1.5 ± 0.2f – – – nd nd – 1.4 ± 0.0f

Values expressed as mean value ± standard deviation (n = 2); significant differences between the values in the same column are indicated by different letters (P b 0.05); G, gelatin based matrix; AA/AP, agar agar/amidated pectin based matrix; EB, elderberry; RC, red cabbage; PC, purple carrot; E, phenolic extract; J, juice; (1), 40 mL (G) and 60 mL (AA/AP) colorant per 200 g basis matrix; (2), 60 mL (G) and 80 mL (AA/AP) colorant per 200 g basis matrix; nd, not determined due to complete color loss. 1 Storage only at 20 °C. 2 Variations due to higher impact of unequally distributed turbidity and air bubbles in uncolored gels, therefore, blank gels were not included in statistical analyses. 3 Value significantly affected by a strong decrease in the L* value.

(PC-E(2)) were stored at 4 °C in the dark and at 25 °C under light exposure, to evaluate the effects of temperature and light on color stability. Since the application of coloring foodstuffs is advantageous regarding regulatory requirements, the RC-J based colorant exhibiting exceptional blue tints was included in this study, despite its less convincing stabilities at 20 °C in the dark (Section 3.4). For this purpose, gels dyed with the latter at both dosages (RC-J(1) and RC-J(2)) were also stored at 4 °C and under illumination at 25 °C. 3.5.1. Cold storage (4 °C) As expected, blue color stability was significantly improved by lowering storage temperature for gels containing RC-J colorant. After storage for 54 days, no color changes were observed in gelatin based gels containing this colorant at both dosages, and AA/AP gels showed only slight fading. Both gel matrices dyed with PC-E(2) did not incur color changes after storage for 5 months (data not shown). Nevertheless, the calculated ΔE* values after 54 days were similar or even higher compared to storage at 20 °C. This discrepancy between the visually assessed stability and calculated ΔE* values were caused by darkening of the gels after 36 days of storage. Lightness increased slightly until day 36, subsequently L* values decreased considerably (G, − 2.5 b ΔL* b − 3.6, Fig. 2; AA/AP, − 2.9 b ΔL* b − 3.0, Fig. 3). Probably, due to a slower decay of the blue ferric anthocyanin chelates, formation of black or gray oxo/hydroxo complexes (Egeter, 1962) may have significantly affected the L* value but not the overall visual appearance. Considering b* values, improved stabilities at 4 °C could be proven. The decrease was significantly slower under cold storage, with Δb* values of 1.0–1.7 and 1.3–2.8 for G (Fig. 2) and AA/AP (Fig. 3) gels, respectively, compared to room temperature. Similar to the trend at 20 °C, color stability was better in gelatin matrix compared to AA/AP dyed with RC-J, and no differences between the two matrices were observed for the PC-E colorant. Astonishingly, blue color stability in gels prepared with PC-E colorant was not improved by cold storage. Visual appearance and Δb* values did not differ when stored at 4 and 20 °C. This finding was in contrast to the previous micro scale study, where model solutions containing PC-E

displayed almost 8 times longer half-life values at 4 °C compared to 20 °C (Buchweitz et al., in press). 3.5.2. Illumination (25 °C + VIS) Light exposure accelerated color fading in gels colored with RC-J. During storage under VIS light at 25 °C, the color changed completely from blue to green and yellowish within 36 days. Analogous to the storage at 4 and 20 °C, stabilities were better in gels based on gelatin, compared to AA/AP, where the blue hue almost faded away after 12 days. As mentioned above for the other storage conditions, at higher colorant dosages fading of the blue color was less pronounced. Very surprisingly, the outstanding stability of the blue color in gels containing the PC-E colorant remained almost unaffected by illumination. After storage under illumination at 25 °C for 54 days, ΔE* values of 1.4 were only marginally higher for both matrices than at 20 °C without exposition to light (ΔE* = 1.1–1.5), hence displaying exceptional light stability of anthocyanin chelates. Presumably, Cyd-3gal-xyl-glc-fer and Cyd-3-gal-xyl-glc-coum amounting to 48.5 mol% and 15.3 mol%, respectively, interacted differently and stronger with ferric ions compared to other anthocyanin structures due to their acylation with hydroxycinnamic acids. However, ferric chelates formed with anthocyanins from red cabbage, also containing acylated anthocyanins, were considerably less stable. Apparently, the position of acylation and the structure of the phenolic acid appeared to be decisive factors for chelate stability. 3.6. Degradation kinetics As reported previously, degradation of the blue color in model solutions followed a pseudo-first-order kinetics (Buchweitz et al., in press; Buchweitz, Carle et al., 2012b). In order to verify this reaction kinetics in gels, the increase of b* values, indicating blue hue loss, was used for kinetic calculations (Eq. (2)). Unexpectedly, and in contrast to our previous study assessing model solutions, blue color loss in gels did not follow first-order kinetics (R 2 ≤ 0.70) when stored at 4 °C. This observation was independent of the colorant type

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

(RC-J(1), RC-J(2), or PC-E(2)) and gel matrix (data not shown). For storage at 20 °C in the dark, color decay followed a first-order kinetics in gelatin and agar–agar based gels dyed with RC-J (R 2 > 0.93) (Table 3). In contrast, data of the color loss of gels dyed with PC-E and EB-E showed moderate (R 2 > 0.73) and poor (R 2 ≤ 0.70, data not shown) fit to the first-order kinetic model, respectively. Under exposition to VIS light at 25 °C, blue color decay followed first-order kinetics in gels dyed with RC-J (0.85 ≤ R 2 ≤ 0.97) but not in gels colored with PC-E(2) (R 2 ≤ 0.3, data not shown). These findings demonstrate the decisive matrix effect of the gels on the mechanism of blue chelate stabilization. Half-life (t1/2) and destruction (D) values were calculated for gels displaying at minimum moderate accordance (R2 >0.73) with a firstorder decay (Table 3). The t1/2 and D values were congruent showing similar tendencies regarding color stability as determined by ΔE* values (see Sections 3.4 and 3.5; Tables 2 and 3). However, t1/2 and D values were overestimated compared to the visual appearance. As an example, calculated D values, reflecting the period required for 90% blue color loss, were 2.4–7.5 months for gels containing RC-J colorant. This was in clear-cut contrast to the green hue observed already after 54 days of storage at 20 °C. Presumably, the exclusive consideration of the decline of negative b* values is inappropriate for describing and predicting blue color loss. 4. Conclusion For the first time, the application of ferric anthocyanin chelates as natural blue colorants was investigated in different gels, suitable for human consumption. For this purpose, the formulation of coloring foodstuffs based on red cabbage juice (RC-J) and coloring extracts from elderberry (EB-E) and purple carrot (PC-E), respectively, was modified to meet the requirements of different food relevant gel matrices. Blue color evolution was evaluated (L*a*b*) depending on colorant dosage, pigment source and gel matrix, and color stability (ΔE*) was determined under varying storage conditions. As expected from previous investigations, outstanding blue hues (gentian blue) were obtained with a food colorant based on RC-J. Nevertheless, depending on gel matrix and colorant dosage, blue hues (cobalt blue) were also achieved by using PC-E and EB-E. Moderate stabilities at room temperature were observed for EB-E and PC-J, however, gelatin matrix and storage at cool temperatures significantly improved color stability, while illumination was particularly detrimental. In contrast, excellent blue color stabilities were observed for PC-E dye, unaffected by the gel matrix and storage conditions (without light exposition at 4 and 20 °C, illumination with VIS light at 25 °C). Depending on colorant and storage conditions, blue color decay in the gels deviated significantly from first-order kinetics,

Table 3 Kinetic parameters of blue color degradation in gels stored at 20 °C and under illumination at 25 °C. Storage conditions

Gel matrix

Colorant

t1/2 [days]

D [months]

R2

20 °C

G

RC-J(1) RC-J(2) PC-E(2) RC-J(1) RC-J(2) PC-E(2) RC-J(1) RC-J(2) RC-J(1) RC-J(2)

49.0 ± 0.7bc 68.1 ± 3.6ab 81.9 ± 2.2a 21.5 ± 0.5de 27.3 ± 4.0cd 73.1 ± 16.9a 10.4 ± 1.0de 16.9 ± 1.2de 2.1 ± 0.0e 5.4 ± 0.2de

5.4 ± 0.1 7.5 ± 0.4 9.1 ± 0.2 2.4 ± 0.1 3.0 ± 0.4 8.1 ± 1.9 1.1 ± 0.1 1.9 ± 0.1 0.2 ± 0.0 0.6 ± 0.0

0.9771 0.9373 0.7335 0.9329 0.9444 0.8638 0.9726 0.9367 0.8507 0.9654

AA/AP

25 °C + VIS

G AA/AP

Mean values ± standard deviation; significant differences are indicated by different letters (P b 0.05); t1/2, half-life value; D, destruction value; gels exhibiting correlation factors R2 ≤ 0.7 are not listed; G, gelatin based gel; AA/AP, agar–agar/amidated pectin based gel; RC-J, red cabbage juice, PC-E, purple carrot extract; (1), lower colorant dosage; (2), higher colorant dosage.

281

thus clearly demonstrating the decisive role of the gel matrix. Satisfactory results for dairy-based gels were only obtained in a gelatin matrix. Contrary to the blue hues obtained in gels containing milk, only violet or pink hues could be achieved with added buttermilk and yoghurt, which was attributed to the presence of high lactic acid contents of fermented dairy products annihilating ferric anthocyanin chelate formation. These results demonstrate the transferability of knowledge from model solutions at a micro scale, obtained in a previous study, to real food matrices. From the present findings it may be assumed that proteins are better suited as food matrices for coloring with anthocyanin chelates than polysaccharides. Despite their excellent blue color retention as well as lowest requirements concerning pH and mineral content, gelatin products, especially based on pork gelatin, are not halal and also inappropriate for vegetarians. To avoid such restrictions, legume- and sunflower-based texturized proteins as recently suggested by Schäfer, Neidhart, and Carle (2011) and Weisz, Schneider, Schweiggert, Kammerer, and Carle (2010), respectively, appear to be promising alternatives. Consequently, further investigations extending the colorant application to a larger spectrum of protein and polysaccharide based hydrocolloids should be undertaken. Acknowledgments One of the authors (M.B.) gratefully acknowledges scholarships from the “Studienstiftung des deutschen Volkes” and the “Schlieben– Lange-program” of Baden-Württemberg. We are grateful to Wild, Germany, and GNT, Germany, for providing the fruit juice concentrates and Herbstreith & Fox, Germany, for the donation of the pectins. Appendix A. Supplementary data Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.foodres.2012.11.030. References Bayer, E., Egeter, H., Fink, A., Nether, K., & Wegmann, K. (1966). Complex formation and flower colors. Angewandte Chemie International Edition, 5(9), 791–797. Buchweitz, M., Nagel, A., Carle, R., & Kammerer, D. R. (2012a). Characterization of sugar beet pectin fractions providing enhanced stability of anthocyanin-based natural blue food colorants. Food Chemistry, 132(4), 1971–1979. Buchweitz, M., Carle, R., & Kammerer, D. R. (2012b). Bathochromic and stabilizing effects of sugar beet pectin and an isolated pectic fraction on anthocyanins exhibiting pyrogallol and catechol moieties. Food Chemistry, 135(4), 3010–3019. Buchweitz, M., Brauch, J., Carle, R., & Kammerer, D. R. (in press). Color and stability assessment of blue ferric anthocyanin chelates in liquid pectin-stabilised model systems. Food Chemistry. http://dx.doi.org/10.1016/j.foodchem.2012.10.090. Castañeda-Ovando, A., Pacheco-Hernández, M. L., Páez-Hernández, M. E., Rodríguez, J., & Galán-Vidal, C. A. (2009). Chemical studies of anthocyanins: A review. Food Chemistry, 113(4), 859–871. Dangles, O., Elhabiri, M., & Brouillard, R. (1994). Kinetic and thermodynamic investigation of the aluminium–anthocyanin complexation in aqueous solution. Journal of the Chemical Society, Perkin Transactions 2(12), 2587–2595. Downham, A., & Collins, P. (2000). Colouring our foods in the last and next millennium. International Journal of Food Science and Technology, 35(1), 5–22. Egeter, H. (1962). Komplexbildung als färberisches Prinzip bei Blütenfarben. Dissertation, Universität Karlsruhe, Germany. Elhabiri, M., Figueiredo, P., Toki, K., Saito, N., & Brouillard, R. (1997). Anthocyanin– aluminium and –gallium complexes in aqueous solution. Journal of the Chemical Society, Perkin Transactions 2, 355–362. Endress, H. U., Mattes, F., & Norz, K. (2009). Pectins. In Y. H. Hui (Ed.), Handbook of food science, technology, and engineering (pp. 140/1–140/35). Boca Raton, FL: CRC Press. Giusti, M. M., & Wrolstad, R. E. (2003). Acylated anthocyanins from edible sources and their applications in food systems. Biochemical Engineering Journal, 14(3), 217–225. Gonnet, J. (1998). Colour effects of co-pigmentation of anthocyanins revisited — 1. A colorimetric definition using the CIELAB scale. Food Chemistry, 63(3), 409–415. Henry, B. S. (1996). Natural food colorants. In G. A. F. Hendry, & J. D. Houghton (Eds.), Natural food colorants (pp. 40–79). Glasgow: Blackie Academic & Professional. Hubbermann, E. M., Heins, A., Stöckmann, H., & Schwarz, K. (2006). Influence of acids, salt, sugars and hydrocolloids on the colour stability of anthocyanin rich black currant and elderberry concentrates. European Food Research and Technology, 223(1), 83–90.

282

M. Buchweitz et al. / Food Research International 51 (2013) 274–282

Jorge, R. A., & Chagas, A. P. (1988). Ion-exchange equilibria between solid aluminum pectinates and Ca, Mn (II), Cu (II) and Fe (III) ions in aqueous solution. Journal of the Chemical Society, Faraday Transactions 1, 84(4), 1065–1073. Maier, T., Fromm, M., Schieber, A., Kammerer, D. R., & Carle, R. (2009). Process and storage stability of anthocyanins and non-anthocyanin phenolics in pectin and gelatin gels enriched with grape pomace extracts. European Food Research and Technology, 229(6), 949–960. Malien-Aubert, C., Dangles, O., & Amiot, M. J. (2001). Color stability of commercial anthocyanin-based extracts in relation to the phenolic composition. Protective effects by intra and intermolecular copigmentation. Journal of Agricultural and Food Chemistry, 49(1), 170–176. McCann, D., Barrett, A., Cooper, A., Crumpler, D., Dalen, L., Grimshaw, K., et al. (2007). Food additives and hyperactive behaviour in 3-year-old and 8/9-year-old children in the community: A randomised, double-blinded, placebo-controlled trial. The Lancet, 370(9598), 1560–1567. Nussinovitch, A. (1997). Agar. Hydrocolloid applications gum technology in the food and other industries (pp. 1–16). London: Blackie Academic & Professional. Reyes, L. F., & Cisneros-Zevallos, L. (2007). Degradation kinetics and colour of anthocyanins in aqueous extracts of purple- and red-flesh potatoes (Solanum tuberosum L.). Food Chemistry, 100, 885–894.

Schäfer, C., Neidhart, S., & Carle, R. (2011). Application and sensory evaluation of enzymatically texturised vegetable proteins in food models. European Food Research and Technology, 232(6), 1043–1056. Souci-Fachman-Kraut online database (n.a). http://www.sfk-online.net accessed October 14, 2011. Stintzing, F. C., & Carle, R. (2004). Functional properties of anthocyanins and betalains in plants, food and in human nutrition. Trends in Food Science & Technology, 15(1), 19–38. Thibault, J. F., & Ralet, M. C. (2003). Physico-chemical properties of pectins in the cell walls and after extraction. In F. Voragen, H. Schols, & R. Visser (Eds.), Advances in pectin and pectinase research (pp. 91–122). Dordrecht: Kluwer Academic Publishers. Weisz, G. M., Schneider, L., Schweiggert, U., Kammerer, D. R., & Carle, R. (2010). Sustainable sunflower processing — I. Development of a process for the adsorptive decolorization of sunflower [Helianthus annuus L.] protein extracts. Innovative Food Science & Emerging Technologies, 11(4), 733–741. Wissgott, U., & Bortlik, K. (1996). Prospects for new natural food colorants. Trends in Food Science & Technology, 7(9), 298–302. Yoshida, K., Mori, M., & Kondo, T. (2009). Blue flower color development by anthocyanins: From chemical structure to cell physiology. Natural Product Reports, 26(7), 884–915.